. The overall objective of this application is to develop new drugs for treatment of mycobacterial infections in AIDS patients, especially M. avium (MAC) and M. tuberculosis (MTB). The target is the mycobacterial cell wall which is quite similar in both pathogens, but which contains a unique chemical structure with no counterpart in mammalian cells. This is a comprehensive program to optimize activity of compounds already developed and pursue emerging leads from more recent synthetic studies. Compounds of the following classes are to be designed and synthesized as potential antimycobacterial agents (1) Oligosaccharides congeneric to the natural substrates of the biosynthetic enzymes that can act as substrate- and bisubstrate-based inhibitors of the biosynthesis of arabinogalactan; (2) Transferase inhibitors based on the natural arabinose donor, beta-D-arabinofuranose-1-monophosphodecaprenol (DPA), a lipid conjugated arabinose donor; and (3) Transition state and bisubstrate analogs that will act as inhibitors of polysaccharide elaboration, particularly mycolylation. Synthetic compounds are to be screened against a panel of MAC and MTB strains.